Adding increasing fractions without averaging numerators

Click For Summary
The discussion revolves around the proposed inequality involving sums of fractions, specifically Theorem 1, which asserts that the product of the average of a series of increasing numerators and the sum of the inverses of a series of increasing denominators is greater than the sum of the fractions themselves. The author provides a proof for the case when n=2 and outlines an inductive approach to generalize the result for larger n. There are inquiries about the validity of the proof, potential simplifications, and whether the inequality has been previously recognized in mathematical literature. Participants note that a related inequality exists and suggest it could serve as a useful exercise for undergraduates. The discussion highlights the complexity of proving inequalities in mathematical analysis.
Afterthought
Messages
29
Reaction score
2
I'm interested in the following inequality (which may or may not be true)

Theorem 1:
##( \sum_{i=1}^n \frac{a_i} {n}\ )( \sum_{i=1}^n \frac{1} {b_i}\ ) > \sum_{i=1}^n \frac{a_i} {b_i}\ ##
Where ##n ≥ 2, a_1 < a_2 < ... < a_n## and ##b_1 < b_2 < ... < b_n##.

My attempt at a proof:

1) When n = 2, this says that ## (\frac{a_1 + a_2} {2})(\frac{1} {b_1}\ + \frac{1} {b_2}) > \frac{a_1} {b_1} + \frac{a_2} {b_2} ##
2) Expanding gives ## \frac{a_1}{2b_1} + \frac{a_2}{2b_1} + \frac{a_1}{2b_2} + \frac{a_2}{2b_2} > \frac{a_1} {b_1} + \frac{a_2} {b_2} ##
3) Because ##a_2 > a_1##, we can replace ##a_2## with ##a_1 + x##, where x is a positive real number, and get ## \frac{a_1}{2b_1} + (\frac{a_1}{2b_1} + \frac{x}{2b_1}) + (\frac{a_2}{2b_2} - \frac{x}{2b_2}) + \frac{a_2}{2b_2} > \frac{a_1} {b_1} + \frac{a_2} {b_2} ##
4) Combining terms and rearranging, we get ##\frac{a_1} {b_1} + \frac{a_2} {b_2} + \frac{x}{2b_1} - \frac{x}{2b_2} > \frac{a_1} {b_1} + \frac{a_2} {b_2} ##
5) Since ## b_2 > b_1, \frac{x}{2b_1} - \frac{x}{2b_2} > 0##, and then the above equality is true.

With a base case of n = 2 proven, I will now use induction to generalize this result:

6) Assume the inequality is true for n-1: ##( \sum_{i=1}^{n-1} \frac{a_i} {n-1}\ )( \sum_{i=1}^{n-1} \frac{1} {b_i}\ ) > \sum_{i=1}^{n-1} \frac{a_i} {b_i}\ ##
7) Setting z = ## \sum_{i=1}^{n-1} a_i\ ##, we can rewrite this as ##( \sum_{i=1}^{n-1} \frac{z} {(n-1)b_i}\ ) > \sum_{i=1}^{n-1} \frac{a_i} {b_i}\ ##
8) Adding ## \frac{a_n} {b_n} ## to both sides gives ##( \sum_{i=1}^{n-1} \frac{z} {(n-1)b_i}\ ) + \frac{a_n} {b_n} > \sum_{i=1}^{n} \frac{a_i} {b_i}\ ##
9) I will now make the following assumption: ## ( \sum_{i=1}^{n} \frac{z + a_n} {nb_i}\ ) > ( \sum_{i=1}^{n-1} \frac{z} {(n-1)b_i}\ ) + \frac{a_n} {b_n} > \sum_{i=1}^{n} \frac{a_i} {b_i}\ ##
By showing that the first term is greater than the second term, we can show that it is also greater than the third, and then we will be done (as the first term is another way of writing the first term in Theorem 1).

The rest is just algebra:

10) Changing denominators: ## ( \sum_{i=1}^{n} \frac{(n-1)(z + a_n)} {n(n-1)b_i}\ ) > ( \sum_{i=1}^{n-1} \frac{nz} {n(n-1)b_i}\ ) + \frac{n(n-1)a_n} {n(n-1)b_n} ##
11) Expanding the first term: ## ( \sum_{i=1}^{n-1} \frac{nz + na_n - z - a_n} {n(n-1)b_i}\ ) + \frac{(n-1)(z + a_n)} {n(n-1)b_n} > ( \sum_{i=1}^{n-1} \frac{nz} {n(n-1)b_i}\ ) + \frac{n(n-1)a_n} {n(n-1)b_n} ##
12) Rearranging: ## ( \sum_{i=1}^{n-1} \frac{nz + na_n - z - a_n} {n(n-1)b_i}\ ) - ( \sum_{i=1}^{n-1} \frac{nz} {n(n-1)b_i}\ ) > \frac{n(n-1)a_n} {n(n-1)b_n} - \frac{(n-1)(z + a_n)} {n(n-1)b_n} ##
13) The above becomes ## ( \sum_{i=1}^{n-1} \frac{na_n - z - a_n} {n(n-1)b_i}\ ) > \frac{na_n - z - a_n} {nb_n} ##
14) Diving both sides by ## na_n - z - a_n ##: ## ( \sum_{i=1}^{n-1} \frac{1} {n(n-1)b_i}\ ) > \frac{1} {nb_n} ##
15) Since ## b_{n-1} > b_{n-2} > ... > b_1 ##, ## ( \sum_{i=1}^{n-1} \frac{1} {n(n-1)b_i}\ ) > \sum_{i=1}^{n-1} \frac{1} {n(n-1)b_{n-1}} = \frac{1} {nb_{n-1}} ##. Similary, ## \frac{1} {nb_{n-1}} > \frac{1} {nb_{n}} ##

Therefore, the inequality in step 14 is true, which means that the inequality in step 9 is true, which proves Theorem 1.
---------

This was my first attempt at a formal proof, so sorry for any mistakes! My questions:

1) Is this a valid proof? Are there any missing steps or mistakes?
2) Is there a simpler proof?
3) Has this inequality been noticed before? Are there any other related stuff I can read up on?

Thanks.
 
Mathematics news on Phys.org
I didn't check each individual expression, but the general idea is right and the key steps look right as well.
If you replace the bi by their inverse (and require that those are in decreasing order), it is a common inequality. A nice homework problem for undergrads I guess.

It looks related to the problem "show that a/b+b/c+c/a >= 3".
 
Here is a little puzzle from the book 100 Geometric Games by Pierre Berloquin. The side of a small square is one meter long and the side of a larger square one and a half meters long. One vertex of the large square is at the center of the small square. The side of the large square cuts two sides of the small square into one- third parts and two-thirds parts. What is the area where the squares overlap?

Similar threads

  • · Replies 1 ·
Replies
1
Views
709
  • · Replies 1 ·
Replies
1
Views
1K
Replies
6
Views
2K
  • · Replies 3 ·
Replies
3
Views
2K
  • · Replies 3 ·
Replies
3
Views
2K
  • · Replies 3 ·
Replies
3
Views
2K
  • · Replies 1 ·
Replies
1
Views
3K
Replies
8
Views
3K
  • · Replies 1 ·
Replies
1
Views
2K
  • · Replies 4 ·
Replies
4
Views
3K